Compute per-cluster nearest-neighbor SNP distances

For every non-singleton cluster, the minimum SNP distance to an isolate in another cluster (min_inter_cluster) and to any isolate outside the cluster (min_inter_isolate). Unlike cluster_pairwise_distances(), this works over the whole clustering at once: it derives each cluster's "other cluster" and "non cluster" comparison sets from the full assignment vector. Singleton clusters are dropped via remove_singleton_clusters(), so their sequences never form a cluster of their own but still count as unclustered isolates toward min_inter_isolate.

Usage

cluster_inter_distances(clusters, snp_dist)

Arguments

clusters

A vector named by sequence IDs giving the cluster each sequence belongs to.

snp_dist

A matrix of SNP distances between isolates. Its row/column names define the full universe of isolates, including those not assigned to any cluster.

Value

A matrix with one row per cluster (named by cluster ID) and columns min_inter_cluster and min_inter_isolate.